Introduction: The main drawback of dental composites – volumetric shrinkage during polymerization, in clinical situations causes stress development at the tooth-restoration interface (so-called polymerization shrinkage stress – PSS) that can compromise restoration longevity. To date, in order to reduce PSS, modifications were proposed in composite chemical structure, insertion technique into the tooth cavity, and photo-activation protocol. By modifying material composition, a new generation of composites was created – bulk-fill, to shorten the working time and reduce PSS. The doctoral dissertation aimed to introduce a novel photo-activation protocol of bulk-fill composites using flexible optical fibers as light guides inserted into the center of the composite restoration, to reduce PSS and directly guide light into deeper layers of the filling. Methods: The research was conducted using a tooth model with a mesial-occlusal-distal cavity, as a mold for the three selected bulk-fill composites: one giomer and two nano-hybrid. The experiment included the use of two light guide types: conventional (rigid glass probe as an integral part of the LED curing unit) and experimental (plastic optical fibers 1 mm or 1.5 mm diameter, connected with the conventional light guide). Thus, two photo-activation protocols were defined: conventional (with the standard light guide placed 1 mm above the filling) and two-step experimental – in the first step using two fibers inserted into the filling; in the second step, the fibers were removed from the filling, the remaining voids were filled with composite, and one final conventional curing cycle was activated. Therefore, the samples were divided into three groups: control (conventional photo-activation protocol) and two experimental (OF1 – using Ï•1 mm optical fibers, and OF1.5 – using Ï•1.5 mm optical fibers). The following non-destructive measurements were carried out: tooth model deformation (as a secondary manifestation of PSS), in real-time by digital holographic interferometry; composite temperature change during polymerization, in real-time by infrared thermography; composite degree of conversion at two time points (immediately after photo-activation and after 24h), by Raman spectroscopy. Results: The impact of the experimental photo-activation protocol in reducing model deformation compared to the conventional one, ranged from 15-35 % lower deformation. The registered reduction was explained by: a lower polymerization rate (indirectly presented by real-time deformation and temperature measurements); lower composite temperature change (18-54 %); and lower degree of conversion immediately after photo-activation (only when the composite received lower radiant exposure compared to the control group or due to reaction acceleration caused by higher irradiance in the first step of photo-activation). For the other examined composites, there was no difference in conversion or higher results were registered compared to the control group. After 24 h, the degree of conversion results showed no significant difference between the experimental and the control groups. Lower conversion values were recorded for the same composites as when immediately measured (with one exception), while higher values were not presented in comparison with the control group. Conclusion: By coupling the optical fibers via the universal plastic connector with the light guide of any LED curing unit, it is feasible to implement the proposed photo-activation protocol into everyday clinical practice. Meanwhile, the application of Ï• 1.5 mm fibers may be preferable, considering that shorter exposure time is required and due to easier filling of the larger voids after fiber extraction from the composite filling. For optimal protocol implementation, it is necessary to know the tip irradiance of the selected light guides.